For the past two years, Joseph Richardson has been trying to figure out how to keep young black men with knife and gunshot wounds from turning up again with similar injuries at Prince George’s Hospital Trauma Center outside Washington, D.C.

Richardson is director of the Violence Intervention Research Project at the trauma center. When these men are admitted, he shows up at their rooms to ask them to take part in his ongoing study on risk factors for repeat violent injuries. Sometimes he finds them handcuffed to a hospital bed, guarded by a police officer or two. Richardson has to walk away. The patients are under arrest and off-limits to him.

Richardson is also a criminologist and associate professor at the University of Maryland. And recently, in the context of a national discussion about police violence, he got to thinking about the lack of access that kept him from asking these men what happened. How many of those handcuffed shooting victims had taken a bullet from a cop, he wondered?

With scant data on how many people are shot by police across the country every year, Richardson sees potential in hospital emergency departments. As a researcher he might not have direct access to patients under arrest, but the doctors and nurses certainly do. He’s proposing that emergency departments step in and capitalize on that unique access to compile an alternative data source.

Doctors And Nurses Could Ask: ‘Who Shot You?’

Richardson views police violence as a public health issue and believes health care providers have a role to play in addressing it. The concept seems simple: At some point during a patient’s visit, emergency department staffers ask patients who shot them, record their answers and report the information to state health departments and the Centers for Disease Control and Prevention.

He’s not suggesting doctors and nurses investigate their patients’ claims, or that this self-reported data would even be completely accurate. After all, in quite a few cases it could be impossible to know who shot you.

Even so, Richardson says that some data are better than none. Hospital-reported numbers along with those recorded by police and media outlets could help define the true scope of police shootings.

In December, around the time Richardson floated his idea in the Journal of Urban Health, the FBI announced plans to expand its database on violent police encounters. For the first time, the agency will collect information on serious injuries, not just fatalities. But it will continue to lean on voluntary reports by local police departments.

Richardson is skeptical that the federal government can solve the data problem. “There has to be a more pioneering, innovative approach to doing it,” he says. That’s what he’s trying to figure out. He notes that information about people who survive police shootings is especially elusive. “The only way we would know that is either the police would have to report that or the hospitals would have to,” he says. “Up to this point, neither entity has done it.”

Richardson points to a 2009 survey of academic emergency physicians that found that almost all of them believed they’d seen cases of excessive use of force by police but had largely failed to report them.

In interviews with the emergency department staff at Prince George’s, he found that the overwhelming majority said the hospital has an ethical responsibility to record and report police-involved shootings. But doctors and nurses raised concerns about the logistics and consequences.

Some said it would be difficult to put into practice a standardized approach to collecting the information. Others felt patients weren’t likely to open up to trauma staff — especially given the presence of police anytime a victim is under arrest. Still others worried they’d be dragged into court to testify if they implicated the police.

Can Hospitals Balance Care And Reporting On Shootings?

Logistics aside, what looms over Richardson’s proposal is a philosophical divide over the role of the hospital and its staff.

As American College of Emergency Physicians board member James Augustine sees it, muddling a hospital’s mission is bad for patients. “The hospital is not a good place for legal and law enforcement activities to infringe on people’s rights for health care,” says the veteran emergency medicine doctor. “In the emergency setting, this is not a priority.”

But he doesn’t dismiss the idea outright. The health care system plays a vital role in amassing data, he says. In fact, many trauma centers already collect reams of information and submit it to the National Trauma Data Bank. Stripped of names, it’s used to track everything from auto accidents to clothing-related burns. It might be feasible to add information about violent police encounters to those data collection efforts, Augustine says.

David Livingston, chief of trauma at University Hospital in Newark, agrees that when it comes to collecting information, hospitals could help. “Emergency departments are the canary in the coal mine of health in our communities,” he says. “They’re a unique public health resource to gather data.”

But there are serious limitations. Two years ago, Livingston and his colleagues analyzed more than 6,000 gunshot wounds treated at his hospital and found that his own trauma unit’s database didn’t account for nearly 20 percent of them. It turned out the emergency department, not trauma, had handled these relatively minor injuries and Livingston and his co-workers only discovered them when they scoured that department’s billing records.

As for Richardson’s proposal, Livingston says it could work in theory. “Is it economically and logistically feasible?” he asks. “We’d like to think it is, but I have my doubts.” Getting detailed information would probably require dedicated staff, he says, and that’s expensive. But he’s quick to point out that similar data on cancer, heart disease, smoking, obesity and other conditions has been collected, with the National Institutes of Health or the National Science Foundation footing the bill. “In that respect,” he says, “Dr. Richardson’s contention to put this on trauma centers is shirking the government’s responsibility.”

Still, Richardson suggests a place to start: hospital-based violence intervention programs. Only about 30 hospitals in the U.S. have these special programs aimed at curbing readmission for violent crimes, but Richardson sees them as prime candidates for pilot projects.

For University of California, San Francisco trauma surgeon Rochelle Dicker, who heads up the violence intervention program at San Francisco General Hospital, keeping tabs on police violence seems like a natural extension of the work her team already does. “Part of our responsibility as physicians is to not just to do the traditional ‘treat and street,’ but to really get to the issues at hand and address violence in a more comprehensive way.” In order to do that, she says, accurate information is key.

“The work is provocative,” she says of Richardson’s proposal, and it will get people talking. “I applaud the author for taking that first step and opening the door.”

Copyright 2016 NPR. To see more, visit http://www.npr.org/.